Top Tweets from Chilean Presidential Candidates

TODO:

• Figure out how to store secrets outside of the code ✅
• Copy code from last script ✅
• Convert output to plotly ✅
• Add xfun::cache_rds to cache tweets and not waste API calls ✅
• Make tweets available as fly-outs ✅
• Try to load the tweet in a box (crosstalk)
• Sentiment analysis of the replies (crosstalk)

library(rtweet)
library(tidyverse)

Registered S3 methods overwritten by 'dbplyr':
  method         from
  print.tbl_lazy     
  print.tbl_sql      
-- Attaching packages ------------------------------------- tidyverse 1.3.1 --
v ggplot2 3.3.5     v purrr   0.3.4
v tibble  3.1.6     v dplyr   1.0.8
v tidyr   1.2.0     v stringr 1.4.0
v readr   2.1.2     v forcats 0.5.1
-- Conflicts ---------------------------------------- tidyverse_conflicts() --
x dplyr::filter()  masks stats::filter()
x purrr::flatten() masks rtweet::flatten()
x dplyr::lag()     masks stats::lag()

library(keyring)
library(plotly)

Registered S3 method overwritten by 'data.table':
  method           from
  print.data.table     

Attaching package: ‘plotly’

The following object is masked from ‘package:ggplot2’:

    last_plot

The following object is masked from ‘package:stats’:

    filter

The following object is masked from ‘package:graphics’:

    layout

# twitter_token <- create_token(
#   app = "pacto-social-1",
#   consumer_key = 
#     keyring::key_get("pacto-social-1-consumer_key"),
#   consumer_secret = 
#     keyring::key_get("pacto-social-1-consumer_secret"),
#   access_token = 
#     keyring::key_get("pacto-social-1-access_token"),
#   access_secret =
#     keyring::key_get("pacto-social-1-access_secret"),
#   set_renv = TRUE)

# Retrieving tweets from Gabriel Boric
tweets_by_boric <-
  xfun::cache_rds({
    get_timeline("@gabrielboric", n = 3200)
  },
  file = "tweets_by_boric.rds")

# Retrieving tweets from Jose Antonio Kast
tweets_by_kast <- xfun::cache_rds({
  get_timeline("@joseantoniokast", n = 3200)
  },
  file = "tweets_by_kast.rds")

# Function to filter organic tweets and rank them by engagement
get_organic_ranked <- function(df) {
  df %>%
    filter(is_retweet == FALSE,
           is.na(reply_to_status_id),
           # only tweets previous to the presidential election
           created_at <= lubridate::ymd(20211220)) %>%
    mutate(rank_engagement = dense_rank(desc(favorite_count + retweet_count))) %>%
    arrange(rank_engagement)
}

Then I apply the function to both dataframes and merge the results with map_df. Next, I filter the result to keep only the 10 more popular tweets of each candidate, and select the relevant columns for the visualization.

tweets_for_plot <-
  list(tweets_by_boric,
       tweets_by_kast) %>%
  map_df(get_organic_ranked) %>%
  filter(rank_engagement <= 10) %>%
  transmute(
    rank_engagement,
    screen_name,
    status_id,
    created_at,
    text,
    favorite_count,
    retweet_count,
    engagement_count = favorite_count + retweet_count,
    status_url
  )

tweets_for_plot

tweets_for_plot2 <- tweets_for_plot %>% 
  mutate(engagement_count = ifelse(screen_name == "gabrielboric",
                                   - engagement_count,
                                   engagement_count))

breaks_values_eng <- pretty(tweets_for_plot2$engagement_count)

abs_k <- abs(breaks_values_eng)/1000

labels_k <- ifelse(abs_k == 0, "0", str_c(abs_k, "K"))

fig <- plot_ly(
  data = tweets_for_plot2,
  x = ~engagement_count,
  y = ~rank_engagement,
  text = ~abs(engagement_count),
  type = 'bar',
  color = ~screen_name,
  colors = c("red", "blue"),
  orientation = 'h',
      hovertemplate = paste0(
      "<b>%{text:,.0f}</b> likes and RTs | ",
      format(tweets_for_plot2$created_at, "%b %d"),
      " | @",
      tweets_for_plot2$screen_name,
      "<br><br>",
      str_wrap(tweets_for_plot2$text),
      "<extra></extra>"
      )
) %>% 
  layout(barmode = 'overlay',
         title = "Most popular tweets by Chilean presidential candidates",
         yaxis = list(title = "Ranking",
                      autorange="reversed",
                      showgrid=T,
                      autotick = F, tickmode = "array", tickvals = 1:10),
         xaxis = list(title = "Engagement (RTs + Likes)",
                      tickmode = 'array',
                      tickvals = breaks_values_eng,
                      ticktext = labels_k),
         legend = list(title=list(text='<b>Candidate</b>'),
                       orientation = "h",   # show entries horizontally
                       xanchor = "center",  # use center of legend as anchor
                       x = 0.5,
                       y=-0.2),
         hovermode = "closest",
         hoverlabel = list(bgcolor = "#DFDFDF", bordercolor = "black")
         )

# TODO: make them be in the same vertical position. ✅
# TODO: Fix the "Rank engagement" labels ✅
# TODO: Rename axis ✅
# TODO: Fix the horizontal axis ("Engagement") ✅
# TODO: Clean flyouts (change sign) ✅
# TODO: add twitter text as flyouts ✅ 
# TODO: change colours ✅

fig

Next

NEXT: crosstalk. https://rstudio.github.io/crosstalk/index.html

Tasks: [✅] Filter tweets after the presidential election [ ] Retrieve the replies of these tweets Possible approach: - Get all the followers of each one - Loop through them and get their timelines - Filter by “in reply to” using the IDs of the tweets I want [ ] Set up a minimal crosstalk example